Sql patch creation impact in busy systems

Hello,

got hit by some all_synonyms bug in PDB 12.1, quick solution seems to be add hint via

sys.dbms_sqldiag_internal.i_create_patch (for 3 sql_ids)

but is it safe in ExaCS environment running hundreds of sessions ?

Should we expect spike not only casued by hard parse for those sql_ids ?

Easy fallback via drop sql patch but same questions remain.

Have not find anything documented how this works .

Regards.

Greg

--
http://www.freelists.org/webpage/oracle-l